Recent Projects

Cupar Analytics has delivered projects across acoustic data processing, ecological modelling, and environmental analysis. Below is a selection of recent work demonstrating end-to-end workflows, from raw data processing through to modelling, interpretation, and operational systems.

Automated detection of fish school in ARIS SONAR swath.

Automated fish monitoring from imaging sonar data

Intro
Development of an end-to-end system for processing ARIS imaging sonar data to estimate fish counts and support monitoring programs.

Main result
A fully operational workflow integrating data processing, machine learning models, uncertainty estimation, and interactive tools. The system produces fish count estimates, length distributions, and activity indices, supported by an application for validation, exploration, and reporting. It is deployed on multi-year datasets and supports routine monitoring and scalable analysis across sites and conditions.

Partners

Fisheries Management Scotland, The Atlantic Salmon Trust, Deveron Bogie & Isla.

Output
Operational monitoring system and reporting outputs for fisheries management.


Forest bat monitoring and population dynamics

Intro
Long-term analysis of forest bat populations within a montane climate refuge, using extensive banding data collected over more than 20 years to estimate population density and understand ecological dynamics.

Main result
Application of spatial capture–recapture (SCR) models to >3,600 capture events across four bat species, producing one of the first robust density estimates for free-ranging forest bats. Densities were higher than previously reported globally and varied by species and elevation, with clear ecological structuring across habitats. Long-term trends revealed substantial interannual variability and evidence of climate-linked responses, including increasing density in one species associated with rising temperatures.

Partners

The Analytical Edge.

Output

B Law, T Brassil, R Proud, J Potts. 2023. Ecology and Evolution 13 (6), e10215.

Large forest bat captured by Dr Brad Law.
Dr Roland Proud and researchers from NaFIRRI (Uganda) performing an echosounder survey in Lake Victoria.

Schistosomiasis, fisheries, and human health — Lake Victoria

Intro
Investigation of the relationship between fisheries, ecosystem dynamics, and human disease risk in Lake Victoria (East Africa).

Main result
Analysis demonstrating how predator stocking (catfish) influences parasite host populations (snails) and human infection risk, linking fisheries management with public health outcomes.

Partners

University of St Andrews, Pelagic Ecology Research Group. Lake Victoria Fisheries Organisation.

Output

Proud et al. 2025. PLOS Neglected Tropical Diseases.

Cloud-based processing of large-scale echosounder datasets 

Intro
Development of a scalable system for processing large acoustic datasets using the OceanBox Data Platform within a Microsoft Azure environment.

Main result
Automated workflows for preparing and analysing large echosounder datasets, enabling efficient data handling, reproducible processing, and integration with downstream analysis tools.

Partners

Sustainovate, OceanBox.

Output
Cloud-based processing workflows and integrated analysis tools.


Inside a Herring school. Echosounders can be used to detect fish schools and other aquatic species.